In recent years, with the in-depth reform of the electricity market and the continuous improvement of the penetration rate of distributed resources in the distribution network, the energy trading between intelligent buildings with dual attributes of production and consumption has brought new opportunities and challenges to the nearby consumption of distributed energy. However, for the microgrid system with multi-intelligent buildings, there are defects such as large amount of communication information, low robustness and user privacy in the process of power trading. At the same time, it will also be affected by various uncertain factors such as the access of new energy and the lack of timeliness of transactions. In order to solve the above problems, this paper proposes a rolling P2P energy trading optimization strategy based on distributed information interaction for multi-intelligent buildings in microgrid.
Firstly, considering the aggregation characteristics of various flexible resources in intelligent buildings, the prediction interval results of distributed photovoltaic power generation and the feasible range of flexible resources are characterized in the form of aggregation power interval by Minkowski summation theory, and the aggregation interval model of P2P transaction is established. Among them, the distributed photovoltaic prediction interval is modeled by transforming the benchmark output at different confidence levels into the prediction quantile for the feasible region. At the same time, an interval rolling P2P energy trading framework is constructed. During the energy management period, each building participates in the rolling P2P energy trading by combining the aggregated power interval with its own electricity purchase and sale strategy. Secondly, the risk cost brought by the uncertainty of photovoltaic output to P2P transactions is quantified by CVaR, and an economic dispatch model with the minimum total operating cost of microgrid multi-intelligent buildings is established. On this basis, the P2P transaction power between buildings is used as a consistency variable, and the P2P transaction power and transaction price are obtained based on the distributed solution of the information interaction between adjacent buildings, and the energy transaction period is continuously pushed backward until it meets the requirements of all intelligent buildings in the microgrid.
In the case analysis, the scheduling results of different buildings in the microgrid and the optimization results of different algorithms are compared respectively, which verifies the effectiveness of the interval rolling P2P energy trading model proposed in this paper. At the same time, the practicability and solution efficiency of the distributed information interaction algorithm in this paper have also been reflected. Through the example analysis, the following conclusions can be drawn: (1) Participating in the energy transaction between buildings in the form of aggregation interval fully taps the scheduling potential of flexible resources in buildings and improves the flexibility of coordinated scheduling of multi-intelligent buildings in microgrid. (2) Compared with the ordinary P2P trading, the rolling P2P energy trading improves the enthusiasm of intelligent buildings to participate in energy trading and the self-consumption level of distributed energy while taking into account the economy of system operation. (3) The distributed information interaction strategy proposed in this paper makes the multi-intelligent buildings in the microgrid only need to interact with the expected transaction volume information, and at the same time solve their own optimization problems in parallel, which has a good fit with the rolling P2P transaction mode. It avoids the problems of high computational pressure and privacy leakage, and improves the convergence speed of distributed information interaction. It has good scalability and can effectively solve the optimization iteration problem of large-scale intelligent buildings.

Electric aircraft has become a major development trend in the future aviation industry due to its advantages of low carbon and environmental protection. The air insulation of electric aircraft needs to withstand high-frequency voltage in high altitude. Therefore, this paper qualitatively studies the air discharge characteristics and microscopic mechanisms between needle-plate electrodes under different pulse voltage parameters and different humidity in the low temperature sub-atmospheric pressure environment of high altitude through simulation and experimentation.
Firstly, the pulse power supply was built by a 4-stage half-bridge Marx circuit. Then, the two-dimensional axisymmetric streamer discharge model of low-temperature sub atmospheric air was built, and three sets of Helmholtz equations were coupled to calculate the photoionization. Finally, the images of air streamer discharge under different conditions were captured by intensified charge coupled device (ICCD).
The following conclusions are drawn through simulation and experiment under the condition of low temperature and sub-atmospheric pressure: (1) The simulation outcomes reveal that when the reduced electric field strength remains the same, as the altitude increases, the breakdown voltage drops, the electron density gradually reduces, the electric field strength of the streamer head decreases, and the development speed of the streamer slows down. As the rising edge of the pulse grows, the electron density decreases simultaneously. When the discharge can be accomplished within one pulse, an increase in the pulse width has minimal effect on the discharge. Under the circumstances of low temperature and sub-atmospheric pressure, with the rise in humidity, the electron density increases concurrently, the peak value of the electric field intensity also rises, and the development speed of the streamer becomes faster. (2) The experimental results indicate that when the reduced electric field strength is consistent, with the increase of altitude, the penetration time of the streamer becomes longer, the channel brightness decreases, and the channel radius increases. When the pulse width of the pulse voltage is greater than the discharge time, the increase in the pulse width has no influence on the discharge process; when the frequency of the pulse voltage rises, the brightness of the streamer channel gradually intensifies; under the condition of low temperature and sub-atmospheric pressure, with the increase in humidity, the penetration time of the streamer becomes shorter and the brightness of the streamer channel increases. (3) Under the same conditions, the simulation and experimental results have a consistent conclusion regarding the development speed of the streamer. The influence of the pulse width on the discharge depends on whether the discharge can be completed within one pulse. The brightness of the streamer channel is positively correlated with the electric field intensity of the streamer head.

The oil-filled terminal adopts a solid-liquid composite insulation composite structure composed of silicone rubber (SiR) stress cone and silicone oil (SO) inside. Compared to the cable body, when the composite insulation interface is invaded by moisture or contains air gaps or impurities, it can cause electric field distortion. This distortion can trigger creepage and even flashover along the surface of the stress cone, significantly impacting the service life of the oil-filled terminal. Among them, moisture intrusion is recognized as the main factor causing insulation deterioration in cable terminals, and cable termination failure rate caused by it account for about 50%. Therefore, the moisture migration process and equilibrium characteristics between SO and SiR solid-liquid medium in the oil-filling terminal need to be further studied. In this paper, the moisture migration law between SO-SiR composite insulation system in oil-filled terminals is systematically studied, the swelling model and mechanism of SO in the terminals are discussed, and the moisture equilibrium characteristics of SO-SiR composite insulation systems under the effect of temperature and swelling are clarified.
Firstly, this study conducted moisture absorption experiments on SO and SiR under various temperature and humidity conditions. The water content of SO and SiR at different temperature and humidity equilibrium states was measured. Using the indirect equilibrium theory, a moisture equilibrium curve for the SO-SiR composite insulation was plotted. The results show that the water content of SO has a linear relationship with the relative humidity at the same temperature, and the saturated water content of SO changes exponentially with temperature. The water content of SiR has a nonlinear relationship with relative humidity, and the saturated water content of SiR does not change with temperature. As the temperature rises, moisture migrates from the SiR to the SO.
In addition to the moisture migration between the SO and the SiR duplex medium in the oil-filled terminal, the SO will also diffuse into the SiR. This diffusion destroys the physical and chemical cross-linking results of the SiR, and affects the moisture absorption characteristics of the SiR. Therefore, it is necessary to clarify the physical mechanism underlying the swelling of SiR by SO. The results show that the SO swells into the SiR in the form of free state and bound state according to the Langmuir diffusion process. With increasing time, the swelling rate increases as a logarithmic function. With increasing temperature, the equilibrium swelling mass remained unchanged, but the swelling rate increased. Under the SO (solvent)-SiR (solute) system, the elastic free energy of the system increased due to the swelling of SO, which was offset by the Gibbs free energy. Finally, the total free energy is zero, and the swelling reaches equilibrium.
On this basis, the moisture equilibrium curve of SO-SiR composite insulation was further optimized. After the swelling of SO, the free volume of SiR increases, which can dissolve more water. However, SiR with different degrees of swelling still exhibits the same water absorption characteristics as unswollen SiR. Combined with the moisture dissolution characteristics of SO, the moisture equilibrium surface diagram of SO-SiR composite insulation under temperature and swelling was drawn. With increased swelling, water molecules migrate from the SO to the SiR. Through this surface diagram, the water content of SO and SiR under different equilibrium states can be obtained, and the operation and maintenance of oil-filled terminals can be guided.

Dry-type transformer is to high voltage level, high power density direction, long-term operation in the electro-thermal cooperative multi-stress complex working conditions such as epoxy resin casting insulation is more likely to induce along the surface flashover failure. In order to study the characteristics of epoxy resin along the surface flashover under the stress of electro-thermal cooperative aging, this paper builds a platform for flashover along the surface under AC stress, and it is found that when the aging temperature is 160℃, the flashover field strength of the epoxy resin specimen aged for 80 days is 2.11 kV/mm, which is a decrease of 25.7%. The steepest decrease in the field strength along the surface is found from 0 to 40 days, which is related to the rapid increase in the surface roughness of the specimen from 0 to 40 days.
A plasma model of epoxy resin flashover along the surface is established by combining the continuity equation of charged particles, the average electron energy equation and the interfacial reaction characterization equation, and the dynamic simulation of flashover along the surface at the working frequency is realized. According to the results of the aging experiment, a random function is introduced to change the surface roughness of the medium, and the dielectric constant after aging is combined to simulate the accurate electro-thermal aging behavior of the epoxy resin, and the temporal and spatial evolution laws of the tangential electric field strength, electron density and surface charge density of the epoxy resin in the process of flashover before and after aging are obtained. The simulation results show that the electron density and surface charge density increase during the flashover development of the aging epoxy resin, and the electric field strength at the head of the flow injection reaches 1.183 kV/mm at 12 ns, an increase of nearly 10.87%. With the aging specimen due to the roughness and dielectric constant increase, its surface charge density will occur surge phenomenon, compared with the aging specimen before the increase of 57.66%, so that the electron density quickly reached the threshold value of the electron collapse to flow injection, resulting in the development of the flashover becomes faster.
The mechanism of combined electro-thermal aging on the surface charge of epoxy resin specimens is explained by the trap effect, and the reason for the decrease in the flash field strength of the specimens is clarified. For the specimen aged for 80 days, the deep trap density and energy level increase to 2.56×10¹⁶ eV^-1·m^-3 and 1.06 eV, respectively, resulting in an increase in the probability of charge entry trapping, which leads to a large amount of surface charge accumulation, and the electric field distortion becomes more serious, thus decreasing the flash-coincidence field strength along the surface. The above findings provide theoretical and methodological basis for the fault operation and maintenance and life prediction of dry-type transformers.

Power battery packs are widely used in new energy electric vehicles and are the core components of electric vehicles. Studying the temperature field modeling of the power battery pack is not only beneficial to understanding its temperature field dynamic characteristics, but is also very important for the structural design and health management of the power battery pack. The temperature field of the power battery pack is described by complex partial differential equations. Since a large number of parameters are unknown and many model parameters show strong time variability, traditional physics-based modeling methods are ineffective in achieving online modeling of the temperature field of the power battery pack. Although methods based on deep learning do not rely on physical models, they require a large amount of experimental data during the training process, the model training time is long, and the real-time performance of temperature field prediction is poor. In response to the above problems, this paper proposes a spatio-temporal modeling of the temperature field of power battery packs based on long short-term memory network.
First, the spatio-temporal separation method is used to extract spatial features and time features under offline conditions. Spatial features are continuously updated with the help of incremental learning, and the long short-term memory (LSTM) network is used to model temporal dynamics. Finally, the updated spatial characteristics and time model are integrated to obtain a prediction model of the power battery pack temperature field.
The proposed method was verified on a power battery pack composed of 24 battery cells. Experimental results show that the proposed method can accurately predict the temperature field of the power battery pack regardless of normal conditions or conditions with air flow interference. Without airflow interference, the single-point temperature prediction error of the proposed method is less than 0.4℃, and the root-mean-square error (RMSE) on the test set is 0.095 1℃. In the presence of airflow interference, the single-point temperature prediction error of the proposed method is less than 0.07℃, and the RMSE on the test set is 0.014 7℃.Under the condition of air flow, the modeling error of the proposed method is smaller. This is because under the condition of air flow interference, the spatial gradient of the temperature change of the power battery pack at the same time is smaller, that is, the temperature change is gentler, making the spatial characteristics of the modeling smoother.
The following conclusions can be drawn from the simulation analysis: (1) the proposed method can accurately predict the temperature field of the power battery pack regardless of normal conditions or conditions with air flow interference. (2) The proposed method can update spatial features in real time through incremental learning, thereby reducing the computational complexity of the method. (3) The proposed method is a purely data-driven method that does not rely on accurate partial differential equations and is therefore suitable for application in temperature field modeling of actual power battery packs.

Under the two-stage voltage control architecture of provincial regulation, the superior dispatching control center directly sends the voltage command value to the automatic voltage control (AVC) sub-stations of each wind farm, and the AVC sub-stations of each wind farm in the wind power cluster independently perform voltage control without communication with each other. In this case, the AVC sub-stations of each wind farm can only obtain the operation data of the local station. The high efficiency and accuracy of reactive power allocation cannot be achieved through AVC master station, which makes the voltage regulation efficiency of wind power cluster low. In addition, due to the different response time of the energy management platform and the wind turbine, the voltage regulation response speed of the wind farm is also different. Wind farms with fast regulation speed bear more reactive power, and wind farms with slow regulation speed bear less reactive power, resulting in unbalanced reactive power and waste of reactive power regulation capacity.
Firstly, this paper analyzes the influence of reactive power regulation period and regulation step of AVC sub-station on the voltage control of wind farm grid-connected point. Considering that the operating parameters of each wind farm equipment in the actual system are relatively fixed, the reactive power regulation period is not easy to change, and the fixed adjustment step cannot take into account the adjustment speed and adjustment accuracy. Therefore, this paper focuses on improving the voltage regulation speed of wind farm by changing the reactive power regulation step length.
Secondly, because the voltage of the wind farm grid-connected point is not only related to the reactive power output of its own station, but also affected by the reactive power output of other stations, this paper proposes a voltage control strategy of the AVC sub-station of wind power plant based on "variable step perturbation observation". This strategy changes the output reactive power of the wind power plant, and then measures the voltage change of the grid-connected point, and evaluate the influence of voltage control of other wind farms on the wind farm grid-connected point, dynamically adjust the reactive power regulation step of AVC sub-station, improve the voltage regulation speed of the wind farm, so that the voltage of the wind farm grid-connected point can enter the voltage dead zone faster.
Thirdly, in order to improve the reactive power imbalance in the wind power cluster, the reactive power constraint relationship of the wind farm stations in the cluster is established by analyzing the voltage reactive power coupling relationship between each wind farm, and considering the difference of the reactive power margin of each wind farm station, the variable step size control strategy is improved, and an improved wind farm voltage control strategy considering reactive power constraint is proposed. The voltage regulation speed and reactive power balance of wind power cluster are considered.
Finally, based on the operating data of a wind power cluster in East China, a simulation model of wind farm convergence system is built to verify the effectiveness of the proposed strategy.

The catenary insulator is a critical component of the traction power supply system for high-speed railways. It not only provides electrical control insulation but also plays an essential role in supporting the catenary arm structure. Therefore, the operational safety of the insulator is directly related to the stability of the entire high-speed railway system. However, the detection of insulator defects is often subject to various interferences due to the complex and dynamic railway environment, resulting in low detection accuracy. Moreover, traditional detection methods generally only identify the presence of defects but fail to provide specific semantic descriptions of these defects. This limitation significantly hampers the efficiency of fault diagnosis and maintenance operations. To address these challenges, this paper proposes a defect description method for insulators based on a diffusion model. This method optimizes existing detection technologies in several ways, enabling the model to not only detect insulator defects more accurately but also generate detailed textual descriptions of these defects.
Firstly, we designed a large-kernel spatial selection feature extraction network. Compared to traditional feature extraction networks, this network captures the feature information of insulator defects through larger spatial convolution kernels, significantly enhancing the model's ability to extract insulator defect features. The model can accurately identify potential defects in the insulator, even in complex backgrounds. Secondly, we proposed a detection decoder with a fusion diffusion mechanism based on the diffusion model. This decoder generates noise boxes and uses inverse Bayesian diffusion to restore predictions of the insulator's true bounding box, significantly improving the model's resistance to background interference. This innovation allows the model to more effectively isolate background noise in complex environments, thereby improving the accuracy of defect detection. Finally, to address the limitations of traditional detection models in semantic description, we designed an encoder and decoder based on a cross-attention mechanism to achieve cross-modal mapping between images and text. By using the BLIP model driven by a text filtering mechanism, the model can generate corresponding textual descriptions of the defects based on the detection results. The functionality not only provides maintenance personnel with more intuitive references but also greatly enhances the efficiency of fault handling. Experimental results validate the effectiveness of our method. The proposed insulator defect detection model achieved the mAP_0.5 of 93.04% and the AR and F1-score of up to 83.22% and 82.91%. The BLEU achieved 83.51%, with CIDEr of 1.94, ROUGE-L of 81.59%, METEOR of 51.50%, and SPICE of 37.88%.
The experimental results lead to the following conclusions: (1) Utilizing a large-kernel spatial selection feature extraction network as the image encoder enhances the insulator defect detection network's ability to focus on key features, thereby improving the model's detection accuracy. (2) To address the issue of insulator defect detection being easily disturbed by complex background environments, a detection decoder with a fusion diffusion mechanism was designed. This decoder performs inverse Bayesian diffusion on the noise boxes generated by the decoder, restoring the prediction of the insulator's true bounding box. The model's ability to resist background interference reduces the loss of semantic information related to insulator defects, and enhances the accuracy of the predicted bounding boxes. (3) A cross-modal mapping module was designed to map the relationship between insulator image defect features and text features. The language modeling encoder outputs a textual description of the insulator defects, completing the detection task. Thus, the proposed model not only offers higher detection accuracy but also generates accurate and detailed semantic descriptions of the defects, meeting the actual needs for insulator defect detection and description.

With the increasing volatility and randomness of uncertain variables such as load and new energy, how to rationally dispatch multiple equipment such as cogeneration, gas boiler and energy storage equipment in integrated energy system (IES) according to the response characteristics of the existing potential response resources in IES to cope with the changes of uncertain variables has become the key to explore the differentiated response ability of IES multiple equipment. To solve the above problems, this paper proposes an economic optimal scheduling method for integrated energy systems, which takes into account variational mode decomposition (VMD) of uncertain variables and green certificate-carbon joint trading. By analyzing and mining the potential differential response ability of multiple devices in IES, the response ability of IES system to uncertainty variable volatility and randomness is improved.
Firstly, according to the commonness and difference of multiple types equipment operating response in time scale and regulatory amplitude in IES, uncertain variables such as wind power, electricity/heat/gas load are decomposed into low/medium/high frequency components with different amplitude and frequency through VMD to adapt to the response characteristics of multiple types equipment. Secondly, on the basis of considering the green certificate trading mechanism (GCT) and the carbon trading mechanism (CET), quantitatively calculate the carbon emission reduction of new energy compared with fossil energy in the process of online access, and offset part of the carbon emission through the carbon emission reduction caused by the green certificate, so that the carbon emission source can be reduced to a certain extent in the calculation of carbon emissions, which indirectly affects the carbon trading mechanism. Based on this, the green certificate-carbon joint trading mechanism is constructed. Finally, the medium and large-sized equipment with large inertia responds to the low-frequency component with low frequency and large amplitude, and the energy storage equipment that needs repeated charging and discharging responds to the medium/high-frequency component with small amplitude and positive/negative periodic oscillation, and then an economic optimisation scheduling model of the IES with the objective of minimising the comprehensive cost is established on the basis of the model, which is then passed layer by layer and iteratedly solved based on the order of VMD's low/medium/high-frequency components.
Through theoretical analysis and case simulation, the following conclusions are drawn: (1) The predicted power, electrical load, thermal load and gas load of wind power are decomposed into low, medium and high frequency components through VMD, which are suitable for the operation characteristics and response characteristics of energy storage equipment. The scheduling method proposed in this paper reduces the operation state of overcharge and overdischarge of energy storage equipment, which can effectively improve the utilization rate of energy storage equipment and further improve the system's ability to absorb new energy. (2) Compared with a single energy storage device, the hybrid energy storage system can better smooth and absorb wind power in different frequency bands according to the different frequency characteristics of wind power, so as to eliminate more wind abandonment and reduce the comprehensive cost of the system. (3) Compared with the single CET or GCT mechanism, the GCT-CET linkage mechanism can not only improve the absorption capacity of renewable energy in the integrated energy system, but also promote the further reduction of carbon emissions of the system.

Sulfur hexafluoride (SF₆), which has strong electronegativity and self-recovery, exhibits excellent insulation and arc-extinguishing capabilities and is widely used in the field of power insulation. However, SF₆ is a strong greenhouse effect gas, and its global warming potential is 23 500 times that of CO₂, and its degradation can significantly reduce the pollution and harm of SF₆ to the atmosphere. Then, there are many kinds of toxic and harmful substances in SF₆ degradation products, among which sulfuryl fluoride (SO₂F₂), as the main decomposition product of SF₆, still has the greenhouse effect and huge toxicity and stable nature. The degradation of SO₂F₂ can improve the harmless degradation process of SF₆ and realize the harmless emission of SF₆. At present, many scholars at home and abroad for the treatment of SO₂F₂ waste gas treatment methods mainly include the alkali treatment method, adsorption method, Non-temperature plasma method, etc., in which the Non-temperature plasma method has the advantages of simple structure, ease of control, high efficiency, etc. Still, there is a problem of poor regulation of the product. By filling the catalyst, the degradation rate can be increased and the product selectivity can be improved. In this paper, the degradation of SO₂F₂ by dielectric barrier discharge (DBD) plasma synergistic filling materials was investigated, and the effects of γ-Al₂O₃, ZSM-5, and glass beads on the degradation of SO₂F₂ with different input powers were investigated.
The experimental platform for SO₂F₂ degradation by DBD plasma synergistic filler materials was first constructed. GC-MS was used to quantify SO₂F₂ and its degradation products, and the SO₂F₂ degradation rate and product content were calculated and detected. The experiments found that the addition of filling materials can improve the discharge conditions of the system, enhancing discharge voltage and current. Furthermore, the filling materials can effectively improve the SO₂F₂ degradation rate and energy efficiency (degradation rate: glass beads>γ-Al₂O₃>ZSM-5>no filler), and also change the decomposition path and product selectivity of SO₂F₂ to produce SO₂ that is easy to handle. 2% SO₂F₂ at a flow rate of 150 mL/min and a power of 100 W. As the input power increases, the degradation rate of SO₂F₂ gradually rises, while the energy efficiency shows an overall decreasing trend. With the filling of glass beads, the degradation rate and energy efficiency of SO₂F₂ were 99.5% and 7.69 g/(kW·h), respectively, and the concentration of SO₂ product was 9 278.56×10^-4%, under the same experimental conditions, the degradation rate of SO₂F₂ was lower than that of γ-Al₂O₃ and glass bead filling when ZSM-5 was filled, but the ZSM-5 filling could make SO₂F₂ decompose completely and directionally to SO₂, at which time the content of SO₂ The SO₂F₂ decomposition products are mainly SO₂, SOF₂, SOF₄ and SiF₄, etc. The results of the study show that the SO₂F₂ degradation rate is lower than that of γ-Al₂O₃ and γ-Al₂O₃ filling, but ZSM-5 filling can almost completely directional decomposition of SO₂F₂ to SO₂, at which time the content of SO₂ is 16 908×10^-4%. The results of the study provide reference solutions for the efficient degradation of SO₂F₂ and the harmless treatment of SF₆. The main decomposition products of SO₂F₂ include SOF₂, SO₂, SOF₄, and OF₂. The addition of a catalyst can alter the decomposition pathway of SO₂F₂, facilitating the generation of the more manageable SO₂. The degradation products also contain a significant amount of SiF₄, indicating that etching reactions have occurred.

As the advancement of Industry 4.0 continues, the power and energy sectors are rapidly undergoing intelligent and digital transformation, leading to the emergence of digital twin technology in the field of electrical equipment. As critical primary equipment, power transformers greatly benefit from the development of digital twin models, which enhance operational reliability, maintenance efficiency, and fault prediction capabilities. However, model-driven digital twin models are often constrained by slow computation speeds. To address this issue, this paper constructs a simplified field-circuit coupled model for oil-immersed power transformers using Modelica, aimed at reducing computational complexity. Additionally, to further enhance computational efficiency, the proper orthogonal decomposition (POD) method is applied to the field computation section for order reduction.
Firstly, we investigate the heat generation, heat dissipation mechanisms, and oil flow circulation of a 35 kV, 800 kV·A scaled-down oil-immersed self-cooled (Oil Natural Air Natural, ONAN) converter transformer prototype. Based on this, a simplified method for coupling thermal and circuit calculations and an equivalent modeling approach for the temperature rise of the converter transformer are proposed. Subsequently, the implementation method and encapsulation form of the thermal circuit coupled model using Modelica are discussed. POD is then employed to reduce the order of the field computation section. Finally, temperature rise experiments on the converter transformer are conducted, and the model's computational data is compared with the experimental results.
The comparison between the model’s computational data and the experimental results reveals significant differences in the range of 0.5 to 2 hours, with the maximum discrepancy reaching 9.8 K at the top sampling point. As the operating time increases, the temperature rise difference gradually diminishes, and the temperatures converge in the steady state. Whether the radiator is considered significantly impacts both the magnitude of the winding temperature rise and the hotspot location. In the steady state, excluding the radiator results in a maximum temperature error of 11.39 K between the model’s calculations and the experimental data, whereas the proposed model's maximum temperature error is 1.37 K, and the full-order model's maximum temperature error is 0.82 K. In terms of computational efficiency, the proposed model takes a total of 3.328 hours under the temperature rise condition, which is 258.65 times faster than the full-order three-dimensional model. Compared to the full-order field-circuit coupled model, the computational speed is increased by 5.1 times.
From the analysis of the model's computational results and the experimental data, the following conclusions can be drawn: (1) The proposed model has a maximum temperature error of 1.37 K compared to the experimental results, making it suitable for temperature rise calculations and winding hotspot analysis of converter transformers. (2) For oil-immersed self-cooled converter transformers, excluding the complete oil flow circulation with the radiator in temperature rise calculations may lead to significant deviations in both the magnitude and location of the winding hotspot temperature rise compared to actual conditions. (3) The proposed model effectively reduces computational costs through the thermal circuit coupling and POD order reduction methods. Compared to the full-order three-dimensional model, the computation speed is increased by 258.65 times, and compared to the full-order field-circuit coupled model, the computation speed is increased by 5.1 times, better meeting the timeliness requirements of digital twin models.